GIS AND MAPPING OF MOVING MARINE SAND DUNES Thierry GARLAN SHOM/HOM - CS 92803 - 29228 BREST Cedex 2-
[email protected]
ABSTRACT: The objective of works on sand dune dynamics, is to predict the evolution of the seafloor and to precisely determine the optimum recurrence for re-surveying these area because mobile dunes represent a serious threat for navigation safety. Horizontal dune displacements could be superior of 20 meters/year and vertical variations could vary by one metre in some weeks. Therefore a project on marine sand dune has been initiated. This paper present a new GIS on marine sand banks and sand dunes of the French continental shelf and a plan for new maps to show these bedforms and their historical movements. Like in deserts and rivers, seas contain dunes. Tide, weather and climate have a decisive role on these bedforms and their movements bring about problems for hydrographic offices because they entail to resurvey sea floor to guarantee safety of navigation. The objectives of our works are to answer to some questions about the time between hydrographic survey and the representation of dunes on charts. In parallel with research on numerical modelling of dune dynamics, we develop a GIS which is the synthesis of knowledge from some decades of surveys. This system is used to compare the knowledge of bedforms with charts, to modify the surveying strategy, to look at the historical movement of the dunes and to develop new cartographical products. By their shape, dunes therefore reflect hydrodynamic conditions. Dunes are sedimentary ripples, often periodical, featuring a crest, a gentle slope and a steep slope. The orientation of their crests is generally perpendicular to the main direction of currents. Marine sand dunes with linear crests have the slowest migration velocity at the opposite, barchan dunes can reach an annual mean displacement up to 70m/y. All dunes do not exhibit a simple morphology, they could be isolated, aggregated and/or superimposed. The most important marine bedforms are sand banks which are more large and more stable, so they are potentially less dangerous, but they often present superimposed dunes. The other two cases are isolated dunes which are more rapid and field of dunes. On charts, it is not easy to see the presence of bedforms. It could be infer by the mariner from the symbol of mobile seabed or from indentations of isobaths. In fact there is a great difference between the knowledge from surveys, and the representation on chart. We created a GIS on sand dunes and sand banks. The principal advantage of it, is to have the possibility of viewing the evolution of the dune along the time. The other progress is to know every parameter of each dune for each step of its evolution. The comparison of movements of dunes with charts is used to decide a new survey and/or a new edition of the chart. The GIS is also used to define the limits and characteristics of dune field along the time. These works concern the safety of navigation but also fishing, sand mining, mine burial, wind farm, ... Charts are not adapted to these needs. We present a master of a new map which show limits of bedforms, the position and height of crests and a vector to show the mean direction and the mean annual speed of movements.
1 INTRODUCTION The ubiquity of dunes is established. Dunes exist in deserts and from rivers to the basis of the continental slope; similar shapes seem detectable in deep ocean environment and on Mars. Their movements bring about problems for hydrographic offices because they entail to resurvey sea floor to guarantee safety of navigation. Tide, weather and climate have a decisive role on sedimentary dynamics. Studying sand dunes should not be limited to instantaneous processes such as swell and tides but should also integrate event-related processes (storms). Thanks to the high resolution of multi-beam echo sounders (MES), it has been possible to increase the morphological accuracy. However the question is for how long (how many week, months or years) can the morphology of a dune be estimated as in conformity with the hydrographic survey? Are charts in agreement with the complex morphology of field of dunes? Can new dunes emerge between two successive hydrographic soundings? If so, for how long can we have the guarantee that no new dune will appear? To answer to these questions, studies and numerical models are done. At the same time, we develop a GIS which is the synthesis of knowledge from some decades of surveys from hydrographic office and research laboratories. This system is used to compare the knowledge of these bedforms with charts, to modify the surveying strategy, to look at the historical movement of the dunes and to develop new cartographical products. 2 AN ASSESSMENT OF THE PRESENT KNOWLEDGE Over two hundred researchers from some thirty countries participated in the Workshops MARID, co organized by the SHOM and which took place in Lille (Trentesaux and Garlan, 2000), Twente (Hulscher, Idier and Garlan, 2004) and Leeds (Parsons, Garlan and Best, 2008). Lectures described case studies, parameterizing of structures, sea and river dune modelling using analytical, digital or physical techniques. Issues concerning the structures and phenomena to be modelled are still at an early stage. What is the impact of grain size? of depth? How do dunes and currents interact? Without answering all these questions, the three workshops were useful to further refine the knowledge and compare problems and progress in the various fields from acquisition to modelling. Why are we only at such an elementary stage in the domain of sub-aqueous sand dune modelling? First, because the localization of sedimentary structures on the sea floor until the eighties was not precise enough to enable scientists to measure displacements of the order of some meters. A second reason is that the data and studies on dunes were, and remain, still too scarce for realistic modelling to be feasible. As Stolk (2000) said "On the migration of sandwaves only a few datasets are available". As far as observations were concerned, Le Bot and al. (2000) observe that gravel in the North Sea dunes are mobilized for several hours at each phase of tidal currents and that the currents generated by medium strength winds and storms modify the asymmetry of tidal currents to the extent they can reverse currents and the direction of dune progress. As remarked by Schüttenhelm (2000): "Sand waves in the southern North Sea are an enigmatic result of a dynamic equilibrium between sand, tidal currents and wave energy." and Mosher and al (2000) add: "The stability of the dune field remains unanswered. Repetitive multibeam bathymetric surveys and long term current flow monitoring are required to answer this
question." For many authors, the sedimentary dynamics of dunes and their modelling are complex because they are poorly defined. Powell and al (2000) consider that "the formation and disappearance of sediment bedforms, occurring under the action of waves and currents, is the result of a complex interaction between the fluid and the underlying sediment; this is poorly understood, yet extremely important, in both physical and numerical models of coastal processes. Bedforms affect bottom roughness and shear stresses, wave attenuation, and sediment transport”. Many questions remain unresolved "What are the physical mechanisms causing sand waves in shallow shelf seas to migrate and at what rate?". Thus these workshops emphasize the deficit of data and the limits of studies which are too local and instantaneous. In 2001, SHOM conducted a survey for monitoring bottom morphology and variations in near-bottom current measurements in the Straits of Dover. This survey showed that the dune short term dynamics can be directly related to the dynamics of small superimposed structures. The displacement velocities recorded, fast for superimposed structures and slow for the dune, indicate that high resolution recurrent surveys are not the best answer to improve hydrographic knowledge. In order to warn navigators against dangerous shoal waters, it would be necessary to measure the surface area occupied by a dune and its peak envelope, and above all to keep a historical record of the dune and hydrodynamic conditions it was submitted to just before the hydrographic survey so as to determine its historical high point, modelling being used to calculate its theoretical height. 3 DUNE SHAPE, NATURE AND MOVEMENTS In order to characterize the dynamics of dunes, high-quality long-term data must be available on the velocity and direction of migration, crest shape and evolution, and grain size of the sediments. Because of the lack of data on dune formation and development, dune formation time is obtained by modelling. Such dune formation models (Blondeaux and al., 2000), still do not provide relevant results in complex sectors such as the North Sea (Idier, 2002). Dunes are sedimentary ripples, often periodical, featuring a crest, a gentle slope and a steep slope. The orientation of their crests is generally perpendicular to the main direction of currents, but angular variations reaching up to 20° are frequently observed (Le Bot, 2001). When the current cannot set in motion all sedimentary particles or when there is a sand deficit, dunes are barchanoid-shaped, whereas they have a linear shape when the current is saturated with sediments. By their shape, dunes therefore reflect hydrodynamic conditions. According to several authors, marine sand dunes with linear crest have the slowest migration velocity (from some meters to 30m/y). On the other hand, barchan dunes are characterized by high velocities, and can reach an annual mean displacement up to 70m/y (Berné and al, 1989). Dune shape is strongly correlated to their nature. Flemming (2000) gives, for different medium grain size, the characteristic values of dune maximum height (Hmax) and wavelength (λ). For instance for very fine sands: Hmax ≈ 0.028 m and λ ≈ 0.14 m, whereas for coarse sediments Hmax ≈ 22.0 m and λ ≈ 380 m.
The compilation of the descriptive parameters of 1 500 dunes from various environments in every seas of the world, enabled Flemming to establish a continuous statistical model ranging from ripples to giant dunes, with the following characteristics: Dune height comprised between 0.001 and 22 meters, Wavelength ranging from 0.01 to 1000 m The equation of the maximum height of dunes is HMax = 0.16 λ 0.84 The classification proposed by Ashley (1990), widely accepted today (see Table 1), concerns the sand structures generated by unidirectional currents, bidirectional currents and the combination of both. The Hmean of the Flemming diagram allows obtaining the height of the corresponding dunes. Such models constitute a very important basis for digital modelling. To give the good answer to the different needs, we split last class in very large dunes (5m
